This is a proof of concept to demonstrate the AI generation of targeted phishing emails by scraping social media profiles.
This software is intended for educational purposes only. The creators and contributors assume no responsibility for its use. By using this software, you acknowledge that all risks and consequences arising from its application are your responsibility.
git clone https://github.com/alvwin/ai_phishing_toolkit
cd ai_phishing_toolkit
python -m venv env
.\env\Scripts\activate.ps1
pip install -r requirements.txt
llama3:latest
and dolphin-mixtral:8x7b
ollama pull llama3:latest
ollama pull dolphin-mixtral:8x7b
To use other models you'd also have to update the code since these two models are hard-coded
li_at
and sessionid
cookiesFor local AI, run ollama serve
Run the program with
python console_program.py
Follow instructions
Run the program with flags
Example:
python console_program.py -ai "OpenAI GPT-3.5" -platform LinkedIn -uname user-name -output SMS -payload "Download Link" -api_key [OpenAI api key]
Available flags:
options:
-h, --help show this help message and exit
-ai {OpenAI GPT-3.5,OpenAI GPT-4,Mistral (local),Llama3 (local)}
Select AI to use
-platform {Twitter,LinkedIn}
Select platform to scrape user data from
-list LIST Target multiple users (provide file path with username list)
-uname UNAME Username of the single user to scrape
-output {Email,SMS,Vishing script,Twitter DM,Twitter post,Twitter reply,Pretext}
Specify what to generate
-payload {Login page,Attachment,Download Link,Else}
Specify type of payload
-template TEMPLATE Specify a template to use
-api_key API_KEY API key for the AI service